uniteTOlove.ca

Shelter Cost Per Bed-Night — Playbook

Shelter costs per bed have climbed for years — separating genuine cost growth from bookkeeping quirks in the numbers.

DRAFTThe playbookThe evidence file

v2.0 · 2026-08-11

---

The honest bottom line

Here's the part that doesn't get said often enough: even counting both real, legitimate factors — system growth and longer average stays — the City's own underlying research says outright that they don't fully explain the cost growth. There's a residual — real, acknowledged, and not yet broken down into what it actually is. Is it staffing costs rising faster than inflation? Richer wraparound services bundled into a "bed-night" now that weren't there in 2010? Something else? Nobody has done that math yet. That's not a gap this document is inventing to sound rigorous — it's the City's own research naming its own biggest unanswered question.

---

a recommendation card — Publish a Current, Itemized Per-Operator Per Diem Schedule

Card id: a recommendation card · Issue: shelter-cost-per-bed-night · Backgrounder: our research file for that page · Trust: carried-forward (CL-548 through CL-554)

Problem

The City's most recent detailed, shelter-by-shelter per diem schedule this library has located dates to 2010 — a full range from $20.25/night (Fife House, no food) to $75.75/night (St. Vincent de Paul – Amelie, transitional) [CL-548 through CL-554]. No comparably itemized current-era (2020s) per-operator schedule has been located despite a dedicated search across this page’s own inherited research and this review's live discovery. The City demonstrably used to publish this level of detail and has not published anything comparably granular since — a declining-transparency finding, not merely an absence of research effort. This card addresses only the publication gap, not any claim about what current rates actually are.

Action

Toronto Shelter and Support Services publishes an annual, itemized, shelter-by-shelter per diem rate schedule for all Purchase-of-Service contracts, in the same format and level of detail as the City's own 2010 schedule.

Jurisdiction split

Cost

Order-of-magnitude: low — a recurring publication/reporting exercise using data the City already collects for contract administration, not a new data system. Comparator: the City already produced exactly this format once, in 2010, indicating the marginal cost of resuming it is a reporting-format decision, not a new capability build [CL-548].

Funding path

Existing TSSS budget-office and contract-administration operating capacity; no new funding mechanism required.

Who benefits, and how

The public, journalists, and this library's own research, via a current, comparable, per-operator cost baseline instead of a schedule frozen at 2010; City Council and TSSS's own budget office, via better data for evaluating cost variation across operators at contract renewal.

Who bears the cost, and how

City taxpayers, negligibly, via existing TSSS administrative staff time; no new payer class identified.

Who benefits from the status quo

No beneficiary identified in the backing backgrounder's Cui Bono table — the backgrounder's own Cui Bono section states this review did not assess whether any operator's funding trajectory meets the Accountability Observatory's provenance bar, and this card does not manufacture a beneficiary the backgrounder didn't establish.

ROI (four dimensions) — schema v2

(a) Financial ROI

Not yet estimable as a dollar figure — a disclosure/reporting change, not a direct cost or saving to the City's own budget; no source prices a fiscal effect for publication itself. Confidence: low.

(b) Economic ROI

Not yet estimable. A recurring data-publication requirement has no direct construction, employment, or induced-spending effect distinct from ordinary administrative reporting, and no comparator was located. Confidence: low.

(c) Social ROI

Directional only: a current, itemized schedule would let researchers, journalists, and Council actually calculate whether the "unexplained residual" in real per-unit cost growth (documented in the backgrounder's own "Key tensions / tradeoffs" as the page’s single largest open analytical task) is concentrated in specific operators or programs, rather than remaining an aggregate, undisaggregated question. No source quantifies a trust or accountability effect from this specific disclosure change. Confidence: low-medium — the mechanism (more granular data enables better analysis) is plausible and consistent with general public-data-transparency literature, but not measured for this specific case.

(d) Environmental ROI

Genuinely environmentally neutral — a data-publication requirement with no construction, land-use, or physical-operations component. Confidence: high.

Evidence

Confidence & uncertainties

Medium-high confidence on the problem statement (the 2010 schedule and its non-repetition since are both directly documented). Low confidence on all four ROI dimensions, each stated as "not yet estimable" or directional-only rather than guessed. This card does not resolve why the City stopped publishing this level of detail — no source in this page’s substrate establishes a reason, and this card does not speculate one.

Status

DRAFT — blocked on: fairness and legal review; confirming whether TSSS already produces this schedule internally without publishing it (this review did not locate one either way).

---

a recommendation card — Disaggregate the Unexplained Cost-Growth Residual Before Its Next Public Citation

Card id: a recommendation card · Card class: INTERNAL · Issue: shelter-cost-per-bed-night · Backgrounder: our research file for that page · Trust: carried-forward (CL-542, CL-543, CL-544, CL-552)

Problem

Toronto's system-wide per diem rate grew from $52.13 (2010) to $136 (2023) — a 161% nominal, ~96% real increase [CL-542, CL-543]. Shelter capacity grew roughly 114-125% and bed turnover has declined every year since 2011 tracking began — both real, legitimate, but only partial explanations for the cost growth [CL-544, CL-552]. The backgrounder's inherited master briefing states directly that "the growth in average length of stay and the growth in system scale do not, on their own, fully explain a near-doubling in real per-unit cost," and names disaggregating that residual into staffing/admin growth, service-intensity growth, and any remaining unexplained portion as its own top unresolved analytical task. This card addresses only that this disaggregation has not been done, not what it would find.

Action

Before the ~96% real per-diem cost-growth figure is next cited in any public-facing document from this library, the citing document either performs or cites an existing disaggregation of that growth into capacity-growth effects, turnover/length-of-stay effects, and a stated remaining residual — rather than repeating the headline figure without that breakdown.

Jurisdiction split

Cost

Order-of-magnitude: low — an analytical exercise using figures already documented in this project's own backgrounder, though it would require sourcing additional data (staffing costs, service-intensity metrics) not yet compiled in this page’s substrate. Comparator: the AG's own budgeted-vs-reported per-bed-night methodology already provides a partial template for how to separate occupancy effects from other cost drivers, though it does not itself disaggregate the multi-year residual this card addresses.

Funding path

Not applicable; this is a documentation/research-quality action internal to this library, not a funded external action.

Who benefits, and how

Readers of this page’s material and any downstream render (day-one paper, brief, journalist briefing) drawing on the ~96% real-growth figure, via an accurate, disaggregated account rather than a headline number that invites either a "waste" or "fully justified" reading neither the evidence nor this page’s own substrate supports.

Who bears the cost, and how

This project's own authoring/editorial capacity, via the work of the disaggregation itself.

Who benefits from the status quo

No beneficiary identified — this is an internal documentation-accuracy question, not a policy question with an external beneficiary of the status quo.

ROI (four dimensions) — schema v2

(a) Financial ROI

Not applicable — a this library's internal records action with no fiscal impact on any government or private entity.

(b) Economic ROI

Not applicable, for the same reason.

(c) Social ROI

Directional: a disaggregated, honest accounting of cost growth strengthens this library's own credibility bar ("would this survive a hostile outside reader checking it directly") more than repeating an unexplained headline percentage would — a research-integrity benefit rather than a measurable public outcome. This dimension does not meaningfully apply to an internal documentation-accuracy card, stated here rather than forced into the four-dimension schema artificially.

(d) Environmental ROI

Not applicable — no physical, construction, or land-use component.

Evidence

Confidence & uncertainties

High confidence that the residual exists and is undisaggregated (both are directly stated in the backgrounder's own inherited master briefing). This card does not itself perform the disaggregation, only names that it should happen before the figure's next public citation — consistent with the backgrounder's own explicit naming of this as its single largest open analytical task.

Status

DRAFT — blocked on: identifying a data source for staffing/admin cost growth and service-intensity change over the 2010-2023 window (neither located in this page’s substrate as of this review) before the disaggregation this card calls for can actually be performed.

---

a recommendation card — Fund the Purpose-Built Shelter Gap Using the City's Own $127/Bed/Night Savings Case

Card id: a recommendation card · Issue: shelter-cost-per-bed-night · Backgrounder: our research file for that page · Trust: carried-forward (CL-544, CL-90655)

Problem

Cost per bed-night varies sharply by program type: base shelters at $136/night, new purpose-built shelters at $126/night, and temporary shelter hotels at $253/night [CL-544]. The $127/night gap between hotels and purpose-built shelters is the City's own stated basis for its 10-year Homelessness Services Capital Infrastructure Strategy (HSCIS) — a $674.5 million capital program of which $507.6 million (13 of 20 planned sites) is funded, leaving $166.9 million and 7 sites unfunded [CL-90655, backgrounder "Capital cost per bed"]. This card addresses the same capital-funding gap the sibling shelter-system-capacity-strain page’s a recommendation card card addresses, from this page’s own cost-per-bed-night evidence base rather than duplicating that card — cited here because the per-bed-night cost gap is this page’s own load-bearing finding, not borrowed without basis.

Action

The City continues pursuing the capital cost-share ask already proposed by the sibling shelter-system-capacity-strain page’s a recommendation card card (provincial Homelessness Prevention Program and federal Reaching Home cost-sharing for the 7 unfunded HSCIS sites), using this page’s own independently-computed $127/bed/night operating-cost differential as additional, corroborating evidence for the capital investment's financial case.

Jurisdiction split

Cost

Order-of-magnitude: the unfunded balance is $166.9 million [CL-90655], identical to the figure a recommendation card addresses — this card does not propose a different dollar figure, only adds this page’s own per-bed-night cost evidence to the existing case.

Funding path

Named, existing mechanisms only: the provincial Homelessness Prevention Program and federal Reaching Home program, identical to a recommendation card's funding path — this card does not invent a new mechanism.

Who benefits, and how

People experiencing homelessness in the 7 unfunded HSCIS sites' catchment areas, via faster access to the $126/night purpose-built model instead of continued reliance on $253/night temporary hotel capacity [CL-544]; City taxpayers, via the operating-cost differential itself once sites open.

Who bears the cost, and how

Under the status quo: City of Toronto taxpayers, via 100% City-financed capital debt (per the sibling backgrounder's own capital-funding figures). Under this card's action: provincial and federal general revenue, sharing a cost currently borne entirely locally.

Who benefits from the status quo

No beneficiary identified in the backing backgrounder's Cui Bono table — the backgrounder's own Cui Bono section states plainly that this review did not assess whether any operator's funding trajectory meets the Accountability Observatory's provenance bar for a graded beneficiary finding.

ROI (four dimensions) — schema v2

(a) Financial ROI

The $127/bed/night differential between hotel ($253) and purpose-built ($126) capacity, independently computed and confirmed in this page’s own backgrounder from City primary figures ($136 base shelter / $126 purpose-built / $253 hotel) [CL-544], corroborates the same $74 million/20-year savings case a recommendation card already cites — this card adds no new dollar figure, only an independently-derived per-bed-night comparator strengthening the evidentiary basis. Confidence: medium — the differential is independently confirmed across two of this project's leaves (this one and shelter-system-capacity-strain), though both ultimately trace to the same City-published figures rather than fully independent sources.

(b) Economic ROI

Not separately modeled in this page; see the sibling shelter-system-capacity-strain page’s a recommendation card card for the closest available comparator (a UK social-housing-construction employment-multiplier study, explicitly flagged there as borrowed cross-country and cross-program-type) — not re-derived independently here, to avoid manufacturing a second, possibly inconsistent estimate for the same gap. Confidence: low.

(c) Social ROI

Directional: purpose-built capacity at the lower per-bed-night cost this page documents plausibly frees up operating budget for service quality or capacity elsewhere in the system, though no claim in this page’s substrate quantifies that reallocation effect [CL-544]. Confidence: low-medium — the mechanism is plausible and grounded in a real, confirmed cost gap, but no source models the downstream reallocation effect.

(d) Environmental ROI

Not separately modeled in this page; see a recommendation card's own modular/prefabrication construction-emissions comparator, which addresses the same physical construction this card's funding ask would enable — not re-derived independently here. Confidence: low-medium, inherited from a recommendation card's own stated confidence.

Evidence

Confidence & uncertainties

Medium confidence. This card deliberately does not duplicate a recommendation card's own ROI modeling wholesale — it states plainly where it corroborates that card's evidence (Financial ROI) versus where it simply inherits the same unresolved gap (Economic/Environmental ROI). CL-90655 remains at “still being checked”, not verified, in the claims register.

Status

DRAFT — blocked on: fairness and legal review; coordination with the sibling page’s a recommendation card card to avoid presenting two independently-worded asks for the same $166.9 million gap as though they were unrelated proposals.

---

Production record

Version: v1.0 (2026-07-14) → v2.0 (playbook pass, 2026-08-11, Lane L3c). Prior classification preserved: a recommendation card carries Card class: INTERNAL (FIX-2, W1b cards audit, 2026-08-06) — its Action names this project's own documentation/research-quality process as the actor, not a public body, per a recorded judgment ruling. Playbook pass: opened with the honest-bottom-line paragraph salvaged from the retired day-one memo (archive/dayone/shelter-cost-per-bed-night.md, now superseded, kept as history); per-card metadata consolidated to one line; empty/thin ROI sub-sections collapsed to one paragraph each; all facts, figures, and claim ids preserved. Status: DRAFT.

About this file